A data processing method, a method for obtaining contribution values, and related equipment.

By directly calculating the contribution value of the display behavior in the recommendation system using two machine learning models, the problem of high computing overhead in the existing technology is solved, and efficient and accurate contribution value calculation is achieved.

CN117112890BActive Publication Date: 2024-11-15HUAWEI TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202310915053.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-07-24
Publication Date
2024-11-15
Estimated Expiration
2043-07-24

AI Technical Summary

Technical Problem

The method of calculating the contribution value of the present behavior in the existing recommendation system requires exhaustive collections of all possible sets, resulting in huge computational overhead and inefficient efficiency.

Method used

By training two machine learning models, the contribution value of each display behavior is directly calculated, avoiding exhaustive traversal of all sets and reducing computing resource consumption.

Benefits of technology

The calculation resource consumption for calculating the contribution value of each display behavior is greatly reduced, the calculation efficiency is improved, and the contribution value obtained is more accurate and in line with the actual situation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117112890B_ABST
    Figure CN117112890B_ABST
Patent Text Reader

Abstract

The embodiments of the present application disclose a data processing method, a contribution value acquisition method and related equipment, and use artificial intelligence technology to attribute display behaviors in a recommendation system. The method includes: processing a training sample through a first machine learning model to obtain first information, the training sample indicates at least one display behavior, and the first information indicates the probability of at least one display behavior being interacted; processing the training sample through a second machine learning model to obtain second information, the second information indicates the sum of the contribution values ​​of at least one display behavior, and the contribution value of each display behavior represents the contribution of each display behavior to at least one display behavior being interacted; training the second machine learning model according to a loss function, the loss function indicates the similarity between the first information and the second information, and the trained second machine learning model is used to obtain the contribution value of each display behavior, thereby reducing the computing resources consumed by the calculation process of "the contribution value of each display behavior".
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular to a data processing method, a contribution value acquisition method, and related equipment. Background Art

[0002] Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a branch of computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that machines have the functions of perception, reasoning and decision-making.

[0003] Applying artificial intelligence technology to recommendation systems is a scenario-based application method. In recommendation systems, the same object can be displayed to users multiple times. Users may interact with the object during a certain display behavior of the object. In order to optimize the delivery strategy of the object, a machine learning model can be used to predict the post-click conversion rate (CVR) of each display behavior. In other words, a machine learning model can be used to predict the contribution value of each display behavior to the interaction with the object.

[0004] A currently adopted method includes: after obtaining n display behaviors of the object and some of the n display behaviors have been interacted, it is necessary to exhaustively enumerate all sets that can be composed of the aforementioned n display behaviors, and use a machine learning model to generate the probability that at least one display behavior in each set is interacted, and then calculate the Sharpe value of each display behavior based on the probability that at least one display behavior in each set is interacted, and determine the Shapley value of each display behavior as the contribution value of each display behavior.

[0005] However, in the process of calculating the contribution value of each display behavior, it is necessary to exhaustively enumerate all sets that can be composed of n display behaviors and calculate the probability that at least one display behavior in each set is interacted, which results in huge computational overhead. Summary of the invention

[0006] The embodiments of the present application provide a data processing method, a contribution value acquisition method, and related equipment. The contribution value of each display behavior in at least one display behavior can be directly obtained through a second machine learning model. There is no need to exhaustively enumerate all sets that can be composed of the aforementioned at least one display behavior, and then traverse the probability of at least one display behavior being interacted in each set, which greatly reduces the computing resources consumed in the calculation process of the "contribution value of each display behavior".

[0007] To solve the above technical problems, the present application provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present application provides a data processing method that can use artificial intelligence technology to attribute display behaviors in a recommendation system. The method includes: a training device inputs a first training sample into a first machine learning model to obtain first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates the probability of at least one first display behavior being interacted with; exemplarily, the first training sample includes descriptive information of each first display behavior.

[0009] The training device processes the first training sample through the second machine learning model to obtain second information, where the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of at least one first display behavior.

[0010] The training device trains the second machine learning model according to the first loss function until the first convergence condition is met, thereby obtaining a second machine learning model that has performed a training operation; wherein the first loss function indicates the similarity between the first information and the second information, and the goal of training using the first loss function includes improving the similarity between the first information corresponding to the first training sample and the second information; the second machine learning model that has performed a training operation is used to obtain a contribution value of each display behavior in at least one display behavior in the application stage.

[0011] In this implementation, the first training sample is input into the first machine learning model to obtain the first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates the probability of at least one first display behavior being interacted; the first training sample is input into the second machine learning model to obtain the second information output by the second machine learning model, wherein the second information indicates the sum of the contribution values ​​of each first display behavior in the at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the at least one first display behavior being interacted; the second machine learning model is trained according to the loss function until the convergence condition is met, and the second machine learning model that has performed the training operation is obtained, and the loss function indicates the similarity between the first information and the second information. The second machine learning model that has performed the training operation can be obtained through the above-mentioned scheme, and the contribution value of each display behavior in the at least one display behavior can be directly obtained through the second machine learning model, and it is no longer necessary to exhaustively enumerate all sets that can be composed of the above-mentioned at least one display behavior, and then traverse the probability of at least one display behavior being interacted in each set, which greatly reduces the computing resources consumed in the calculation process of "the contribution value of each display behavior".

[0012] In one possible implementation, at least one first display behavior indicated by the first training sample may be a display behavior for the same object. For example, at least one first display behavior pointed to by the first training sample may be a display behavior for the same application. For another example, at least one first display behavior pointed to by the first training sample may be a display behavior for the same news. For another example, at least one first display behavior pointed to by the first training sample may be a display behavior for the same product, etc.

[0013] In one possible implementation, the target display behavior is any one of at least one first display behavior, wherein the contribution value of the target display behavior is obtained based on a first value and a second value corresponding to the target display behavior, the first value corresponding to the target display behavior represents the contribution value of the target display behavior itself, and the second value corresponding to the target display behavior represents the influence value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample on the contribution value of the target display behavior.

[0014] In this implementation, in the process of generating the contribution value of the target display behavior, the impact values ​​of other display behaviors on the contribution value of the target display behavior are also considered. The aforementioned scheme can reflect the actual situation when the user sees an object multiple times, that is, the aforementioned scheme is more consistent with the actual situation, which is conducive to obtaining a more accurate contribution value for each display behavior.

[0015] In one possible implementation, the first training sample includes the display time of each first display behavior, wherein a contribution value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample to the target display behavior is related to a first time distance, and the first time distance is the distance between the display time of the target display behavior and the display time of other first display behaviors.

[0016] In this implementation, the longer the time between two advertisements seen by the user, the smaller the impact of the previous advertisement on the current advertisement will be, and the contribution value of other display behaviors to the target display behavior in this application is also related to the time distance between the two display behaviors. That is, the method for determining the contribution value of the target display behavior in this application is more consistent with the actual situation, which is conducive to improving the accuracy of the contribution value of each display behavior finally obtained.

[0017] In one possible implementation, a first parameter is used when calculating the influence value of a first display behavior whose display time is earlier than a target display behavior in at least one first display behavior on a contribution value of the target display behavior, and a second parameter is used when calculating the influence value of a first display behavior whose display time is later than the target display behavior in at least one first display behavior on a contribution value of the target display behavior, and the first parameter and the second parameter are not completely the same.

[0018] In one possible implementation, the first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

[0019] In this implementation, after updating at least one display time in the third training sample, a second training sample is obtained, and the second training sample is also placed in the training data set of the second machine learning model, so that the training data set of the second machine learning model is enhanced, that is, the training samples of the second machine learning model are enriched, which is conducive to making the training process of the second machine learning model more sufficient, so as to improve the accuracy of the information output by the trained second machine learning model.

[0020] In one possible implementation, the value of the parameter in the second machine learning model that has been trained is used to obtain the contribution value of each display behavior in at least one display behavior in the application stage. In this implementation, the training device only needs to send the value of at least one parameter in the second machine learning model that has been trained to the execution device, and the execution device can obtain the third machine learning model based on the value of the aforementioned at least one parameter, which greatly reduces the communication resources consumed by the "deployment process of the third machine learning model" and improves the convenience of the "deployment process of the third machine learning model".

[0021] In one possible implementation, the second machine learning model includes a first module and a second module, the first module is used to generate a contribution value of each display behavior in at least one display behavior, and the second module is used to perform weighted summation on the contribution value of each display behavior in at least one display behavior generated by the first module to obtain a total contribution value corresponding to at least one display behavior; "the second machine learning model that has performed a training operation is used to obtain the contribution value of each display behavior in at least one display behavior" can be understood as: the first module in the second machine learning model after performing a training operation is used to obtain the contribution value of each display behavior in at least one display behavior in the application stage.

[0022] In one possible implementation, the first machine learning model is pre-trained, and the first machine learning model is any of the following: a recurrent neural network, a convolutional neural network, a fully connected neural network, or an attention-based neural network. In this implementation, since the first machine learning model is pre-trained, that is, when the training of the second machine learning model begins, the accuracy of the first information output by the first machine learning model is relatively high, and the first information output by the first machine learning model is used for distillation learning of the second machine learning model, then the first machine learning model is pre-trained, which is conducive to improving the speed of the training phase of the second machine learning model, and is also conducive to making the information output by the trained second machine learning model more accurate; in addition, the method provided in this application is compatible with the first machine learning model using various types of neural networks, which is conducive to expanding the application scenarios of this solution and greatly improving the implementation flexibility of this solution.

[0023] In a possible implementation, the probability of at least one first display behavior being interacted is any of the following: the probability of being clicked in at least one first display behavior, the probability of being downloaded in at least one first display behavior, the probability of being purchased in at least one first display behavior, or the probability of being played in at least one first display behavior. In this implementation, multiple specific implementation schemes of "at least one first display behavior being interacted" are provided, which expands the application scenarios of this scheme and improves the implementation flexibility of this scheme.

[0024] In a second aspect, an embodiment of the present application provides a method for obtaining a contribution value, which can use artificial intelligence technology to attribute display behaviors in a recommendation system. The method includes: an execution device obtains a first sample, the first sample indicates at least one display behavior, and at least one display behavior has been interacted with; exemplarily, the first sample includes description information of each display behavior in the aforementioned at least one display behavior.

[0025] The execution device generates a contribution value of each display behavior pointed to by the first sample through a third machine learning model, and the contribution value of each display behavior represents the contribution of each display behavior to at least one display behavior being interacted with, and the third machine learning model is obtained based on the second machine learning model that has performed a training operation. Among them, the first loss function used in the training operation indicates the similarity between the first information corresponding to the first training sample and the second information, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of at least one first display behavior being interacted with, the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to at least one first display behavior being interacted with.

[0026] In one possible implementation, the third display behavior is any one of at least one display behavior indicated by the first sample, wherein the contribution value of the third display behavior is obtained based on a first value and a second value corresponding to the third display behavior, the first value corresponding to the third display behavior represents the contribution value of the third display behavior itself, and the second value corresponding to the third display behavior represents the influence value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample on the contribution value of the third display behavior.

[0027] In one possible implementation, the first sample also includes the display time of the display behavior indicated by the first sample, wherein a contribution value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample to the third display behavior is related to a second time distance, and the second time distance is the distance between the display time of the third display behavior and the display time of other display behaviors.

[0028] In one possible implementation, the first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

[0029] In one possible implementation, “the third machine learning model is obtained based on the second machine learning model that has performed a training operation” can be understood as: the third machine learning model is obtained based on the value of at least one parameter in the second machine learning model that has performed a training operation.

[0030] In one possible implementation, the second machine learning model includes a first module and a second module, the first module is used to generate a contribution value of each display behavior in at least one display behavior, and the second module is used to perform weighted summation on the contribution value of each display behavior in at least one display behavior generated by the first module to obtain a total contribution value corresponding to at least one display behavior; "the third machine learning model is obtained based on the second machine learning model that has performed a training operation" can be understood as: the third machine learning model is the first module in the second machine learning model after the training operation has been performed.

[0031] The meanings of the terms in the second aspect and various possible implementations of the second aspect and the beneficial effects brought about can all be found in the first aspect and will not be repeated here.

[0032] In a third aspect, an embodiment of the present application provides a data processing device that can use artificial intelligence technology to attribute display behaviors in a recommendation system, and the device includes: a processing module, which is used to input a first training sample into a first machine learning model to obtain first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates the probability of at least one first display behavior being interacted; a processing module, which is used to process the first training sample through a second machine learning model to obtain second information, and the second information indicates the sum of contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of at least one first display behavior; a training module, which is used to train the second machine learning model according to a loss function until a convergence condition is met, and obtain a second machine learning model that has performed a training operation, wherein the loss function indicates the similarity between the first information and the second information, and the second machine learning model that has performed a training operation is used to obtain the contribution value of each display behavior in at least one display behavior.

[0033] In the third aspect of the present application, the data processing device is also used to execute the steps performed by the training device in the first aspect and various possible implementation methods of the first aspect. The specific implementation methods of the steps in the third aspect, the meanings of the terms and the beneficial effects brought about can all be referred to the first aspect and will not be repeated here.

[0034] In a fourth aspect, an embodiment of the present application provides a contribution value acquisition device, which can use artificial intelligence technology to attribute display behaviors in a recommendation system, and the device includes: an acquisition module, used to acquire a first sample, the first sample indicates at least one display behavior, and at least one display behavior has been interacted; a generation module, used to generate a contribution value of each display behavior pointed to by the first sample through a third machine learning model, the contribution value of each display behavior represents the contribution of each display behavior to the at least one display behavior being interacted, and the third machine learning model is obtained based on a second machine learning model that has performed a training operation; wherein the loss function used in the training operation indicates the similarity between the first information and the second information corresponding to the first training sample, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of at least one first display behavior being interacted, the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the at least one first display behavior being interacted.

[0035] In the fourth aspect of the present application, the contribution value acquisition device is also used to execute the steps performed by the execution device in the second aspect and various possible implementation methods of the second aspect. The specific implementation methods of the steps in the fourth aspect, the meanings of the terms and the beneficial effects brought about can all be referred to the second aspect and will not be repeated here.

[0036] In the fifth aspect, an embodiment of the present application provides an execution device, including a processor and a memory, wherein the processor is coupled to the memory, the memory is used to store programs; the processor is used to execute the programs in the memory, so that the execution device executes the contribution value acquisition method described in the second aspect above.

[0037] In a sixth aspect, an embodiment of the present application provides a training device, including a processor and a memory, wherein the processor is coupled to the memory, the memory is used to store programs; the processor is used to execute the programs in the memory, so that the training device executes the data processing method described in the first aspect above.

[0038] In a seventh aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer-readable storage medium is run on a computer, the computer executes the method described in the first aspect or the second aspect above.

[0039] In an eighth aspect, an embodiment of the present application provides a computer program product, which includes a program. When the program runs on a computer, the computer executes the method described in the first or second aspect above.

[0040] In a ninth aspect, the present application provides a chip system, which includes a processor for supporting the implementation of the functions involved in the above aspects, for example, sending or processing the data and / or information involved in the above methods. In one possible design, the chip system also includes a memory, which is used to store program instructions and data necessary for a terminal device or a communication device. The chip system can be composed of a chip, or it can include a chip and other discrete devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1a A schematic diagram of the structure of the artificial intelligence main framework provided in the embodiment of the present application;

[0042] Figure 1b A schematic diagram of a recommendation interface provided in an embodiment of the present application;

[0043] Figure 1c Another schematic diagram of a recommendation interface provided in an embodiment of the present application;

[0044] Figure 2 A system architecture diagram of a data processing system provided in an embodiment of the present application;

[0045] Figure 3 A flowchart of a data processing method provided in an embodiment of the present application;

[0046] Figure 4 Another flowchart of the data processing method provided in the embodiment of the present application;

[0047] Figure 5 A schematic diagram of obtaining a second training sample based on a third training sample provided in an embodiment of the present application;

[0048] Figure 6 Another flowchart of the data processing method provided in the embodiment of the present application;

[0049] Figure 7 A schematic diagram of a process for obtaining a contribution value provided in an embodiment of the present application;

[0050] Figure 8 A schematic diagram of the structure of a data processing device provided in an embodiment of the present application;

[0051] Fig. 9 A schematic diagram of a structure of a contribution value acquisition device provided in an embodiment of the present application;

[0052] Fig.10 A schematic diagram of the structure of an execution device provided in an embodiment of the present application;

[0053] Fig.11A schematic diagram of the structure of a training device provided in an embodiment of the present application;

[0054] Fig.12 A schematic diagram of the structure of a chip provided in an embodiment of the present application. DETAILED DESCRIPTION

[0055] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0056] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0057] First, the overall workflow of the artificial intelligence system is described. Figure 1a , Figure 1a The figure shows a structural diagram of the main framework of artificial intelligence. The following is an explanation of the above artificial intelligence theme framework from the two dimensions of "intelligent information chain" (horizontal axis) and "IT value chain" (vertical axis). Among them, the "intelligent information chain" reflects a series of processes from data acquisition to processing. For example, it can be a general process of intelligent information perception, intelligent information representation and formation, intelligent reasoning, intelligent decision-making, intelligent execution and output. In this process, the data has undergone a condensation process of "data-information-knowledge-wisdom". The "IT value chain" reflects the value that artificial intelligence brings to the information technology industry from the underlying infrastructure of human intelligence, information (providing and processing technology implementation) to the industrial ecology process of the system.

[0058] (1) Infrastructure

[0059] The infrastructure provides computing power support for the artificial intelligence system, enables communication with the outside world, and is supported by the basic platform. It communicates with the outside world through sensors; computing power is provided by smart chips, which can specifically use hardware acceleration chips such as central processing units (CPU), embedded neural network processing units (NPU), graphics processing units (GPU), application specific integrated circuits (ASIC) or field programmable gate arrays (FPGA); the basic platform includes distributed computing frameworks and networks and other related platform guarantees and support, which can include cloud storage and computing, interconnected networks, etc. For example, sensors communicate with the outside world to obtain data, and these data are provided to the smart chips in the distributed computing system provided by the basic platform for calculation.

[0060] (2) Data

[0061] The data on the upper layer of the infrastructure is used to represent the data sources in the field of artificial intelligence. The data involves graphics, images, voice, text, and IoT data of traditional devices, including business data of existing systems and perception data such as force, displacement, liquid level, temperature, and humidity.

[0062] (3) Data processing

[0063] Data processing usually includes data training, machine learning, deep learning, search, reasoning, decision-making and other methods.

[0064] Among them, machine learning and deep learning can symbolize and formalize data for intelligent information modeling, extraction, preprocessing, and training.

[0065] Reasoning refers to the process of simulating human intelligent reasoning in computers or intelligent systems, using formalized information to perform machine thinking and solve problems based on reasoning control strategies. Typical functions are search and matching.

[0066] Decision-making refers to the process of making decisions after intelligent information is reasoned, usually providing functions such as classification, sorting, and prediction.

[0067] (4) General ability

[0068] After the data has undergone the data processing mentioned above, some general capabilities can be further formed based on the results of the data processing, such as an algorithm or a general system, for example, translation, text analysis, computer vision processing, speech recognition, image recognition, etc.

[0069] (5) Smart products and industry applications

[0070] Smart products and industry applications refer to the products and applications of artificial intelligence systems in various fields. They are the encapsulation of the overall artificial intelligence solution, which productizes intelligent information decision-making and realizes practical applications. Its application areas mainly include: smart terminals, smart manufacturing, smart transportation, smart homes, smart medical care, smart security, autonomous driving, smart cities, etc.

[0071] The method provided by the present application can be applied to recommendation systems in various application fields. Specifically, it can be used to attribute display behaviors in the recommendation system. For example, after displaying an object to a user once or multiple times, at least one display behavior of the object can be obtained. When there is an interactive display behavior in the aforementioned at least one display behavior, the aforementioned at least one display behavior can be attributed to optimize the subsequent delivery strategy of the object. "Attributing display behaviors" can be understood as determining the contribution of each display behavior to at least one display behavior being interacted with, that is, allocating the credit of at least one display behavior being interacted with to each display behavior.

[0072] For example, the objects displayed to the user may be any of the following: an application (APP), news, audio, video, product description information, a purchase link or other types of objects, etc., and the interactive behaviors performed by the user may be any of the following: clicking, downloading, purchasing or other interactive behaviors, etc., which are not limited in the embodiments of the present application.

[0073] Since the method provided by the present application can be used in smart terminals, smart homes or other application fields, the application scenarios in multiple application fields of the present application are exemplified below.

[0074] Application field 1: Intelligent terminal field

[0075] For example, see Figure 1b , Figure 1b A schematic diagram of a recommendation interface provided in an embodiment of the present application. Mobile phones, tablets, laptops, or other smart terminals can be deployed Figure 1b The app store shown in Figure 1b The application store shown in the figure can display multiple applications to the user. Figure 1b The interface shown shows the same application to the user (eg Figure 1bAfter the application 1 in the example is viewed multiple times, the user's installation operation on the application 1 is obtained; the multiple display behaviors of the application 1 can be attributed, that is, the contribution of each display behavior of the application 1 to the installation of the application 1 can be determined, so as to optimize the subsequent delivery strategy of the application 1.

[0076] For another example, see Figure 1c , Figure 1c Another schematic diagram of the recommendation interface provided in the embodiment of the present application. Figure 1c The news recommendation interface shown in Figure 1c After the interface shown shows the same news (e.g., News 2) to the user multiple times, the user's click operation on News 2 is obtained; the multiple display behaviors of News 2 can be attributed, that is, the contribution of each display behavior of News 2 to the click of News 2 can be determined to optimize the subsequent delivery strategy of News 2. Recommendation interfaces of other types of objects can also be deployed on smart terminals, such as audio recommendation interfaces, commodity recommendation interfaces, etc. The examples here are only for the convenience of understanding this solution and are not used to limit this solution.

[0077] Application field 2: Smart home field

[0078] Exemplarily, a display screen may be deployed on the refrigerator in a smart home, through which recipes and purchase links for the ingredients involved in the recipes are displayed to users. After the purchase link for the same ingredient (e.g., celery) is displayed to users multiple times in the aforementioned manner, the user's purchase operations on the purchase link for celery are obtained, and the multiple display behaviors of the purchase link for celery can be attributed. That is, the contribution of each display behavior of the purchase link for celery to the user's purchase operation can be determined, so as to optimize the subsequent delivery strategy for the purchase link for celery.

[0079] It should be noted that the method provided in the present application can also be applied to other application scenarios. The above examples of various application scenarios of the present application are only for the convenience of understanding the present solution and are not used to limit the present solution.

[0080] In the above application scenarios, when at least one display behavior of an object is interacted, the contribution value of each display behavior in the at least one display behavior can be determined by the method provided by this application. The "contribution value of each display behavior" represents the contribution of each display behavior to the interaction of at least one display behavior. Before describing the method provided by this application in detail, please refer to Figure 2 , Figure 2 A system architecture diagram of a data processing system provided in an embodiment of the present application, in Figure 2In the embodiment, the data processing system 200 includes a training device 210 , a database 220 , an execution device 230 and a data storage system 240 , and the execution device 230 includes a computing module 231 .

[0081] The database 220 stores a training data set. In the training phase, the training device 210 can generate a second machine learning model 201, and iteratively train the second machine learning model 201 using the training data set to obtain a second machine learning model 201 that has been trained. The second machine learning model 201 can be specifically expressed as a neural network or a non-neural network model. For details, please refer to Figure 3 , Figure 3 A flowchart of a data processing method provided in an embodiment of the present application.

[0082] 301. The training device 210 inputs a first training sample into a first machine learning model to obtain first information output by the first machine learning model, where the first training sample indicates at least one first display behavior, and the first information indicates a probability of at least one first display behavior being interacted with.

[0083] Optionally, the first machine learning model may be a machine learning model that has been trained. Exemplarily, the first information may indicate the probability of being clicked, downloaded, purchased, played, or performing other types of interactive behaviors in at least one first display behavior.

[0084] 302. The training device 210 inputs the first training sample into the second machine learning model 201 to obtain second information output by the second machine learning model 201, where the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of at least one first display behavior.

[0085] 303. The training device 210 iteratively trains the second machine learning model 201 according to the first loss function until the first convergence condition is met, thereby obtaining a second machine learning model that has been trained, wherein the first loss function indicates the similarity between the first information and the second information, and the second machine learning model 201 that has been trained is used to obtain a contribution value of each display behavior in at least one display behavior.

[0086] In an embodiment of the present application, the above-mentioned scheme can be used to obtain a second machine learning model 201 that has performed a training operation. The second machine learning model 201 can directly obtain the contribution value of each display behavior in at least one display behavior. There is no need to exhaustively enumerate all sets that can be composed of the aforementioned at least one display behavior, and then traverse the probability of at least one display behavior being interacted in each set, which greatly reduces the computing resources consumed in the calculation process of the "contribution value of each display behavior".

[0087] A third machine learning model is deployed in the computing module 231 of the execution device 230, and the third machine learning model is obtained based on the second machine learning model 201 that has performed a training operation. In the application stage, the execution device 230 can obtain a first sample, the first sample indicates at least one display behavior, and at least one display behavior has been interacted with; the execution device 230 generates a contribution value of each display behavior pointed to by the aforementioned first sample through the third machine learning model.

[0088] In some embodiments of this application, please refer to Figure 2 , the execution device 230 and the client device may be integrated into the same device, and the user may directly interact with the execution device 230. Exemplarily, when the client device is a desktop computer, the execution device 230 may be a module in the host CPU of the desktop computer that uses the first machine learning model to process data, and the execution device 230 may also be a neural network processor (NPU) in the desktop computer, and the NPU is mounted on the host processor as a coprocessor, and the host processor assigns tasks.

[0089] It is worth noting that Figure 2 It is only a schematic diagram of the architecture of two data processing systems provided by an embodiment of the present invention, and the positional relationship between the devices, components, modules, etc. shown in the figure does not constitute any limitation. For example, in other embodiments of the present application, the execution device 230 and the client device can be independent devices, and the execution device 230 is configured with an input / output (I / O) interface to interact with the client device for data. After determining the first sample, the client device sends the first sample to the execution device 230 through the I / O interface. After the execution device 230 generates the contribution value of each display behavior pointed to by the first sample through the first machine learning model 201 in the calculation module 231, the contribution value of each display behavior pointed to by the aforementioned first sample can be returned to the client device through the I / O interface.

[0090] In combination with the above description, the specific implementation process of the training phase and the application phase of the method provided in the embodiment of the present application will be described below.

[0091] 1. Training Phase

[0092] In the embodiment of the present application, the training phase describes the process in which the training device 210 trains the second machine learning model 201 using the training data set in the database 220. For details, please refer to Figure 4 , Figure 4 Another flow chart of the data processing method provided in the embodiment of the present application is as follows. The data processing method provided in the embodiment of the present application may include:

[0093] 401. Input the fourth training sample into the first machine learning model to obtain first information corresponding to the fourth training sample output by the first machine learning model, the fourth training sample indicates at least one fourth display behavior, and the first information corresponding to the fourth training sample indicates the probability of the aforementioned at least one fourth display behavior being interacted.

[0094] In the embodiment of the present application, step 401 is an optional step. The fourth training sample refers to a training sample used when training the first machine learning model, and the fourth training sample indicates at least one fourth display behavior, that is, the fourth training sample may include description information of each fourth display behavior in at least one fourth display behavior.

[0095] Exemplarily, the aforementioned at least one fourth display behavior may be a display behavior for the same object. For example, the at least one fourth display behavior pointed to by the fourth training sample may be a display behavior for the same application; for another example, the at least one display behavior pointed to by the fourth training sample may be a display behavior for the same news; for another example, the at least one display behavior pointed to by the fourth training sample may be a display behavior for the same product, etc. The examples here are only for the convenience of understanding this solution and are not used to limit this solution.

[0096] Optionally, the at least one display behavior pointed to by the fourth training sample may include at least one display behavior of the same object to the same user; for example, the at least one display behavior may be the behavior of displaying the same object to the same user at different times.

[0097] Each fourth training sample at least includes the display time of each fourth display behavior in at least one fourth display behavior; illustratively, the display time of each fourth display behavior can adopt the timestamp of each fourth display behavior. Optionally, the fourth training sample may also include the display location of each fourth display behavior; for example, the aforementioned display location may be home, company, on the way to work, on the way home from get off work, or other, etc.; for another example, the aforementioned display location may also be the address where the fourth display behavior is performed, etc., which may be determined in combination with actual conditions. Alternatively, the fourth training sample may also include other descriptive information of each fourth display behavior, etc. It should be noted that the examples given here are only for the convenience of understanding this solution and are not used to limit this solution.

[0098] Exemplarily, each fourth training sample can be represented as a sequence [a1, a2, ..., a N ], each element in the above sequence includes a description of the fourth display behavior, and each element in the above sequence can be expressed as a vector [e1, e2, …, e N ], the above vector [e1,e2,…,e t ] can carry a timestamp of a fourth display behavior.

[0099] It should be noted that the concepts of the two terms "second training sample" and "third training sample" will be used in subsequent steps, and the meanings of the two terms "second training sample" and "third training sample" will not be introduced here for the time being.

[0100] Exemplarily, the first machine learning model can be any of the following: a recurrent neural network (RNN), a convolutional neural network (CNN), a fully connected neural network, a neural network based on an attention mechanism or other types of neural networks, etc. The method provided in the present application is compatible with the first machine learning model using various types of neural networks, which is conducive to expanding the application scenarios of the present solution and greatly improving the implementation flexibility of the present solution.

[0101] The first information corresponding to the fourth training sample indicates the probability of at least one fourth display behavior being interacted. Exemplarily, the first information corresponding to the fourth training sample indicates the probability of the last fourth display behavior being interacted among at least one fourth display behavior. The aforementioned "interaction" may refer to any of the following: click, download, purchase, play or other types of interactive behaviors; for example, if the object targeted by the fourth display behavior is an application, the aforementioned "interaction" may be click or download; for another example, if the object targeted by the fourth display behavior is news, audio or video, the aforementioned "interaction" may be click or play; for another example, if the object targeted by the fourth display behavior is a commodity, the aforementioned "interaction" may be click or purchase, etc. The specific definition of "interaction" can be flexibly determined in combination with the actual application scenario, and is not limited in the embodiments of the present application. In this implementation method, a variety of specific implementation schemes of "at least one display behavior being interacted" are provided, which expands the application scenarios of this scheme and improves the implementation flexibility of this scheme.

[0102] 402. Train the first machine learning model according to a second loss function, where the second loss function indicates a similarity between the first information corresponding to the fourth training sample and the expected information.

[0103] In the embodiment of the present application, step 402 is an optional step. The expected information corresponding to the fourth training sample indicates whether at least one fourth display behavior pointed to by the fourth training sample is interacted, that is, the expected information corresponding to the fourth training sample indicates whether the last fourth display behavior displayed in the at least one fourth display behavior pointed to by the fourth training sample is interacted.

[0104] After acquiring the first information corresponding to the fourth training sample, the first training device can generate a function value of the second loss function according to the first information corresponding to the fourth training sample and the expected information corresponding to the fourth training sample, perform gradient derivative on the function value of the second loss function, and use the back propagation algorithm to update the parameters in the first machine learning model, thereby completing a training of the first machine learning model. The second loss function indicates the similarity between the first information corresponding to the fourth training sample and the expected information, and the goal of iterative training using the second loss function includes improving the similarity between the first information corresponding to the fourth training sample and the expected information.

[0105] Exemplarily, the second loss function may be a cross entropy loss function, an L1 loss function, an L2 loss function, or other types of loss functions, etc. The specific type of loss function may be flexibly determined in combination with the actual situation and is not limited here. To further understand this solution, an example of the second loss function is disclosed as follows:

[0106]

[0107] in, represents the second loss function, y represents the expected information corresponding to the fourth training sample, that is, whether the last fourth display behavior displayed in the at least one fourth display behavior pointed to by the fourth training sample is interacted; The first information corresponding to the fourth training sample output by the first machine learning model indicates the probability of the last fourth display behavior displayed among the at least one fourth display behavior pointed to by the fourth training sample being interacted with.

[0108] The first training device repeatedly performs steps 401 and 402 to implement iterative training of the first machine learning model until the second convergence condition is met, thereby obtaining the first machine learning model that has been trained. For example, the second convergence condition may be a convergence condition that satisfies the second loss function, or the number of iterative trainings of the first machine learning model reaches the first number.

[0109] 403. Obtain a first training sample from the training data set, where the first training sample indicates at least one first display behavior.

[0110] In an embodiment of the present application, a training data set may be deployed in the second training device, and the training data set includes multiple training samples. When the second training device trains the first machine learning model, it is necessary to obtain one or more first training samples from the training data set, and the first training sample indicates at least one first display behavior; wherein the meaning of "first training sample" is similar to the meaning of "fourth training sample", and the difference is that the fourth training sample is a training sample used when training the first machine learning model, and the first training sample is a training sample used when training the second machine learning model.

[0111] The second training device and the first training device can be the same device, or the second training device and the first training device can be different devices. After obtaining the first machine learning model that has performed the training operation, the first training device can send the first machine learning model that has performed the training operation to the second training device.

[0112] In one case, there may be a second training sample among multiple training samples in the training data set, wherein the second training sample is obtained based on a third training sample, the third training sample indicates at least one second display behavior, and the third training sample includes description information of at least one second display behavior, and the description information of each second display behavior includes at least the display time of each second display behavior; optionally, the description information of each second display behavior may also include the display location of each second display behavior, or other types of description information, etc.

[0113] The second training sample is obtained by updating at least one display time in the third training sample, that is, at least one display behavior pointed to by the second training sample in the training data set may be a display behavior that has not been actually performed on the user, and the second display behavior pointed to by the third training sample in the training data set may be a display behavior that has actually been performed on the user.

[0114] It should be noted that the specific forms of expression of the "second training sample" and the "third training sample" are similar to the specific forms of expression of the "first training sample", that is, the specific forms of expression of the "second training sample" and the "third training sample" are similar to the specific forms of expression of the "fourth training sample", and will not be repeated here. The "first training sample" obtained from the training data set can be a "second training sample", that is, at least one display behavior pointed to by the first training sample may contain a display behavior that has not been actually performed on the user; or, the aforementioned "first training sample" may also be the second training sample. Similarly, the aforementioned "first training sample" may be a "third training sample" or may not be a "third training sample".

[0115] Regarding the specific implementation process of the second training device generating the second training sample based on the third training sample, exemplarily, the second training device can obtain the third training sample from the original training data set, and each display behavior pointed to by each training sample in the original training data set can be a display behavior that has been performed to the user. Here, the third training sample is s=[(a1, t1), (a2, t2),…, (a N ,t N )] as an example, where a1, a2, …, a N represents the identification information of the N display behaviors pointed to by the third training sample, t1 represents the timestamp of a second display behavior identified by a1, t2 represents the timestamp of a second display behavior identified by a2, and so on. N Represents a N The timestamp of the second display behavior, from t1 to t N The timestamps represented are getting later and later.

[0116] The second training device can be a1, a2, ..., a N The order of is randomly shuffled to get the shuffled sequence In order from past to future (that is, from morning to night), Generate a timestamp, that is, update at least one presentation time in the third training sample, and obtain the following second training sample:

[0117]

[0118] The second training device repeats the above steps for multiple times to generate multiple second training samples, and incorporates each second training sample into the original training data set to achieve data enhancement for the original training data set to obtain an updated training data set.

[0119] It should be noted that the second training device may also use other methods to achieve “obtaining the second training sample based on the third training sample”, for example, not using a1, a2, …, a N The order of a1, a2, ..., a N The timestamp corresponding to any one or more identification information is modified, etc. The above examples are only used to prove the feasibility of this solution and are not used to limit this solution.

[0120] For a more intuitive understanding of this solution, please refer to Figure 5 , Figure 5 A schematic diagram of obtaining a second training sample based on a third training sample provided in an embodiment of the present application, such as Figure 5As shown, the third training sample includes description information of multiple display behaviors, and the description information of each display behavior includes identification information of each display behavior and display time (ie Figure 5 In order to obtain a second training sample based on the third training sample, the second training device can randomly shuffle the identification information of the N display behaviors in the third training sample to obtain Figure 5 The identification information of display behavior N, the identification information of display behavior 3, ..., the identification information of display behavior 2.

[0121] The second training device generates a new presentation time for each identification information of the display behavior in the aforementioned “identification information of display behavior N, identification information of display behavior 3, ..., identification information of display behavior 2”, that is, Figure 5 In the “timestamp A, timestamp C, ..., timestamp P”, the time represented by the aforementioned “timestamp A, timestamp C, ..., timestamp P” is getting later and later. It should be understood that Figure 5 The examples are only for facilitating the understanding of this solution and are not intended to limit this solution.

[0122] In an embodiment of the present application, a second training sample is obtained after updating at least one display time in the third training sample, and the second training sample is also placed in the training data set of the second machine learning model, so that the training data set of the second machine learning model is enhanced, that is, the training samples of the second machine learning model are enriched, which is conducive to making the training process of the second machine learning model more sufficient, so as to improve the accuracy of the information output by the trained second machine learning model.

[0123] In another case, multiple training samples in the training data set may all be collected based on actual display behaviors to users, that is, at least one display behavior pointed to by each training sample in the training data set is a display behavior actually performed to users.

[0124] 404. Input the first training sample into the first machine learning model to obtain first information corresponding to the first training sample output by the first machine learning model, where the first information corresponding to the first training sample indicates a probability of at least one first display behavior being interacted with.

[0125] In the embodiment of the present application, the first information corresponding to the first training sample indicates the probability of at least one first display behavior being interacted, that is, the first information corresponding to the first training sample indicates the probability of at least one first display behavior displayed last in the first display behaviors being interacted. The concept of "first information corresponding to the first training sample" is similar to the concept of "first information corresponding to the fourth training sample" described above, the difference being that "fourth training sample" is replaced by "first training sample", and "fourth display behavior" is replaced by "first display behavior", and the detailed concept of "first information corresponding to the first training sample" is not repeated here.

[0126] Steps 401 and 402 are both optional steps. If steps 401 and 402 are performed, the first machine learning model in step 404 is pre-trained. If steps 401 and 402 are not performed, step 403 may be performed directly, and the first machine learning model may be an initialized machine learning model.

[0127] 405. Process the first training sample through the second machine learning model to obtain second information, where the second information indicates the sum of contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of at least one first display behavior.

[0128] In an embodiment of the present application, the second training device inputs the first training sample into the second machine learning model, processes the first training sample through the second machine learning model, and obtains second information output by the second machine learning model.

[0129] Exemplarily, the second training device processes the first training sample through the second machine learning model, which may include: the second training device first generates a contribution value for each first display behavior through the second machine learning model, and then generates the sum of the contribution values ​​of all first display behaviors in at least one first display behavior through the second machine learning model.

[0130] For the convenience of description, any one of the at least one first display behavior is referred to as the target display behavior in the following. Optionally, the contribution value of the target display behavior can be obtained based on the first value corresponding to the target display behavior and at least one second value, the first value corresponding to the target display behavior represents the contribution value of the target display behavior itself, and each of the at least one second value corresponding to the target display behavior represents the influence value of each other first display behavior other than the target display behavior in the at least one first display behavior indicated by the first training sample on the contribution value of the target display behavior.

[0131] Further optionally, a contribution value of each first display behavior other than the target display behavior in at least one first display behavior indicated by the first training sample to the target display behavior is related to a first time distance, and the first time distance is the distance between the display time of the target display behavior and the display time of the aforementioned other first display behaviors.

[0132] Optionally, a first parameter is used when calculating the influence value of a first display behavior whose display time is earlier than a target display behavior in at least one first display behavior on the contribution value of the target display behavior, and a second parameter is used when calculating the influence value of a first display behavior whose display time is later than the target display behavior in at least one first display behavior on the contribution value of the target display behavior, and the first parameter and the second parameter are not completely the same.

[0133] For example, if the second machine learning model is specifically expressed as a neural network, the second machine learning model can be any one of the following: a recurrent neural network, a convolutional neural network, a neural network based on an attention mechanism, a fully connected neural network or other types of neural networks, etc., which is not limited in the embodiments of the present application.

[0134] Alternatively, the second machine learning model may also be expressed as a non-neural network model. To further understand the present solution, an example in which the second machine learning model is expressed as a non-neural network model is disclosed as follows:

[0135]

[0136] in, represents the second information obtained based on the first training sample, that is, the sum of the contribution values ​​of each first display behavior in the at least one first display behavior pointed to by the first training sample, and the target display behavior is the i-th first display behavior in the at least one first display behavior pointed to by the first training sample; represents the sum of the contribution values ​​of the i-th first display behavior to the last first display behavior in the at least one first display behavior. When the value of j is equal to i, it is w i,i (that is, the impact value of the ith first display behavior on the contribution value of the ith first display behavior), which can also be understood as the contribution value of the ith first display behavior itself, represents the sum of the influence values ​​of the contribution values ​​of the 1st to the i-1th first display behaviors in the at least one first display behavior to the i-th first display behavior, t i represents the display time of the i-th first display behavior, t j represents the display time of the jth first display behavior, α and β are parameters that need to be iteratively updated during the training process of the second machine learning model. It should be understood that the example in formula (2) is only for the convenience of understanding this solution and is not used to limit this solution.

[0137] In an embodiment of the present application, in the process of generating the contribution value of the target display behavior, the impact values ​​of other display behaviors on the contribution value of the target display behavior are also considered. The aforementioned scheme can reflect the actual situation when the user sees an object multiple times, that is, the aforementioned scheme is more consistent with the actual situation, which is conducive to obtaining a more accurate contribution value for each display behavior.

[0138] The longer the time between two advertisements seen by the user, the smaller the impact of the previous advertisement on the current advertisement will be, and the contribution value of other display behaviors to the target display behavior in this application is also related to the time distance between the two display behaviors. That is, the method for determining the contribution value of the target display behavior in this application is more consistent with the actual situation, which is conducive to improving the accuracy of the contribution value of each display behavior finally obtained.

[0139] 406. Train the second machine learning model according to the first loss function until the first convergence condition is met, so as to obtain a second machine learning model that has been trained, wherein the first loss function indicates the similarity between the first information and the second information corresponding to the first training sample, and the second machine learning model that has been trained is used to obtain a contribution value of each display behavior in at least one display behavior.

[0140] In an embodiment of the present application, after obtaining the first information corresponding to the first training sample and the second information corresponding to the first training sample, the second training device can generate a function value of the first loss function based on the first information and the second information corresponding to the first training sample, derive the function value of the first loss function, and use the back propagation algorithm to update the parameter values ​​of the second machine learning model to achieve one training of the second machine learning model.

[0141] The first loss function value indicates the similarity between the first information corresponding to the first training sample and the second information, and the goal of training using the first loss function includes improving the similarity between the first information corresponding to the first training sample and the second information. For example, the first loss function can adopt a cross entropy loss function, an L1 loss function, an L2 loss function, or other types of loss functions, etc., which are not limited in the embodiments of the present application.

[0142] To further understand the present solution, an example of the first loss function is disclosed as follows:

[0143]

[0144] Among them, L a represents the first loss function, represents first information corresponding to the first training sample generated by the first machine learning model, Represents the second information corresponding to the first training sample generated by the second machine learning model. It should be understood that the example in formula (3) is only for the convenience of understanding this solution and is not used to limit this solution.

[0145] Exemplarily, in one implementation, if steps 401 and 402 are performed, the first machine learning model is pre-trained, and the second training device training the second machine learning model according to the first loss function may include: the second training device adjusting the parameters of the second machine learning model according to the first loss function while keeping the parameters of the first machine learning model unchanged. In the embodiment of the present application, since the first machine learning model is pre-trained, that is, when the training of the second machine learning model begins, the accuracy of the first information output by the first machine learning model is relatively high, and the first information output by the first machine learning model is used for distillation learning of the second machine learning model, then pre-training the first machine learning model is beneficial to improving the speed of the training phase of the second machine learning model, and is also beneficial to making the information output by the trained second machine learning model more accurate.

[0146] In another implementation, if steps 401 and 402 are not performed, the second training device training the second machine learning model according to the first loss function may include: the second training device training the first machine learning model and the second machine learning model according to the first loss function and the third loss function. The third loss function indicates the similarity between the first information corresponding to the first training sample and the expected information corresponding to the first training sample, and the expected information corresponding to the first training sample indicates whether at least one first display behavior pointed to by the first training sample is interacted, that is, the expected information corresponding to the first training sample indicates whether the last displayed first display behavior among the at least one first display behavior pointed to by the first training sample is interacted.

[0147] Exemplarily, the second training device generates a function value of the first loss function and a function value of the second loss function according to the first information, the second information and the expected information corresponding to the first training sample, and updates the parameters of the first machine learning model and the second machine learning model based on the back propagation algorithm according to the function value of the first loss function and the function value of the second loss function to achieve one training of the first machine learning model and the second machine learning model.

[0148] It should be noted that the meanings of the "third loss function" and the "second loss function" are similar. The difference is that the "fourth training sample" in the second loss function is replaced by the "first training sample" in the third loss function, and the "fourth display behavior" is replaced by the "first display behavior". The implementation details of the third loss function will not be repeated here.

[0149] The training device repeatedly executes the above steps to implement iterative training of the second machine learning model until the first convergence condition is met, thereby obtaining a second machine learning model that has performed the training operation.

[0150] Exemplarily, in one case, if steps 401 and 402 are performed, the first convergence condition may be that the convergence condition of the first loss function is satisfied, and / or the number of times the second machine learning model is trained reaches a preset number of times. In another case, if steps 401 and 402 are not performed, the first convergence condition may be that the convergence conditions of the first loss function and the third loss function are satisfied, and / or the number of times the second machine learning model is trained reaches a preset number of times.

[0151] For a more intuitive understanding of this solution, please refer to Figure 6 , Figure 6 Another flow chart of the data processing method provided in the embodiment of the present application is as follows: Figure 6 As shown, the first machine learning model is iteratively trained using the training samples in the original training data set until the second convergence condition is met, thereby obtaining a trained first machine learning model, wherein each training sample in the original training data set indicates at least one display behavior; the specific implementation process of the aforementioned steps can be found in the description of steps 401 and 402, which will not be elaborated here.

[0152] After data enhancement is performed on the original training data set, an updated training data set is obtained. The specific implementation process of the above steps can refer to the description in the above step 403, which will not be elaborated here.

[0153] Obtain a first training sample from the updated training data set, input the first training sample into the trained first machine learning model and the second machine learning model respectively, and iteratively train the second machine learning model according to the first information corresponding to the first training sample generated by the trained first machine learning model and the second information corresponding to the first training sample generated by the second machine learning model until the first convergence condition is met, thereby obtaining a second machine learning model that has been trained. The specific implementation process of the aforementioned steps can be referred to the description in steps 404 to 406, which will not be elaborated here. Among them, the second machine learning model that has been trained is used to obtain the contribution value of each display behavior in at least one display behavior in the application stage. It should be understood that Figure 6 The examples are only for facilitating the understanding of this solution and are not intended to limit this solution.

[0154] In one case, "the second machine learning model that has been trained is used to obtain the contribution value of each display behavior in at least one display behavior" can be understood as: the value of at least one parameter in the second machine learning model that has been trained is used to obtain the contribution value of each display behavior in at least one display behavior. Taking the above formula (2) as an example, the values ​​of α and β in formula (2) are used to calculate the contribution value of each display behavior in at least one display behavior. The training device only needs to send the value of at least one parameter in the second machine learning model that has been trained to the execution device, and the execution device can obtain the third machine learning model based on the value of the aforementioned at least one parameter, which greatly reduces the communication resources consumed by the "deployment process of the third machine learning model" and improves the convenience of the "deployment process of the third machine learning model".

[0155] In another case, the second machine learning model includes a first module and a second module, the first module is used to generate a contribution value of each display behavior in at least one display behavior, and the second module is used to perform weighted summation on the contribution value of each display behavior in at least one display behavior generated by the first module to obtain a total contribution value corresponding to at least one display behavior; "the second machine learning model that has performed a training operation is used to obtain the contribution value of each display behavior in at least one display behavior" can be understood as: the first module in the second machine learning model after performing a training operation is used to obtain the contribution value of each display behavior in at least one display behavior.

[0156] 2. Application Phase

[0157] In the embodiment of the present application, the application stage describes the process in which the execution device 230 processes the first sample through the third machine learning model. For details, please refer to Figure 7 , Figure 7 A flow chart of a method for obtaining a contribution value provided in an embodiment of the present application is provided. The method for obtaining a contribution value provided in an embodiment of the present application may include:

[0158] 701. Obtain a first sample, where the first sample indicates at least one display behavior, and the at least one display behavior has been interacted with.

[0159] In an embodiment of the present application, the execution device may obtain a first sample, where the first sample includes description information of each display behavior in the at least one display behavior; illustratively, the description information of each display behavior may include the display time of each display behavior.

[0160] Optionally, the description information of each display behavior may also include the display location of each display behavior or other information, etc., which is not limited in the embodiments of the present application. It should be understood that the specific expression form of the "first sample" is similar to the specific expression form of the "fourth training sample", except that the "fourth display behavior" in the fourth training sample is replaced by the "display behavior" in the first sample, and the specific expression form of the first sample is not repeated here.

[0161] 702. Generate a contribution value for each display behavior pointed to by the first sample through a third machine learning model, the contribution value of each display behavior represents the contribution of each display behavior to at least one display behavior being interacted with, and the third machine learning model is obtained based on a second machine learning model that has performed a training operation; wherein, the first loss function used in the aforementioned training operation indicates the similarity between the first information corresponding to the first training sample and the second information, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of at least one first display behavior being interacted with, the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to at least one first display behavior being interacted with.

[0162] In one embodiment of the present application, "the third machine learning model is obtained based on the second machine learning model that has been trained" can be understood as: the third machine learning model is obtained based on the value of at least one parameter in the second machine learning model that has been trained. Figure 4 Taking formula (2) shown in the corresponding embodiment as an example, the third machine learning model can be obtained based on the values ​​of α and β in the second machine learning model that has performed a training operation.

[0163] To further understand the present solution, the formula used when processing the first sample using the third machine learning model is disclosed as follows:

[0164]

[0165] The first sample indicates T display behaviors, w i Represents the contribution value of the ith display behavior among the T display behaviors pointed to by the first sample, that is, represents the contribution made by the ith display behavior to the interaction of T display behaviors. Represents the sum of the contribution values ​​of the 1st to the i-1th display behaviors in the first sample to the i-th display behavior, represents the sum of the contribution values ​​of the ith display behavior to the ith display behavior in the first sample. When the value of j is equal to i, it represents the contribution value of the ith display behavior itself among the T display behaviors. i represents the display time of the i-th first display behavior, t j represents the display time of the jth first display behavior. The values ​​of α and β are obtained based on the trained second machine learning model. It should be understood that the example in formula (4) is only for the convenience of understanding this solution and is not used to limit this solution.

[0166] In another case, the second machine learning model includes a first module and a second module, the first module is used to generate a contribution value of each display behavior in at least one display behavior, and the second module is used to perform weighted summation on the contribution value of each display behavior in at least one display behavior generated by the first module to obtain a total contribution value corresponding to at least one display behavior; "the third machine learning model is obtained based on the second machine learning model that has performed a training operation" can be understood as: the third machine learning model is the first module in the second machine learning model after the training operation has been performed.

[0167] For example, after obtaining the first sample, the execution device may process the first sample through the third machine learning model to obtain the contribution value of each display behavior pointed to by the first sample; the third machine learning model is obtained based on the second machine learning model that has performed the training operation, and the training process of the second machine learning model can refer to the above Figure 4 The description in the corresponding embodiments will not be repeated here.

[0168] In order to have a more intuitive understanding of the beneficial effects brought by the embodiments of the present application, experiments were conducted on the Criteo dataset. The beneficial effects brought by the present application are demonstrated below in combination with the experimental data.

[0169] Table 1

[0170] AUC Basic 0.9744 Basic+Attention 0.9742 Basic+Shapley 0.9747 The method provided by this application 0.9753

[0171] Among them, Basic, Basic+Attention and Basic+Shapley are two existing methods for generating the contribution value of each display behavior. AUC is an evaluation index of "the contribution value of each display behavior generated by the machine learning model". The value of AUC is between 0.5 and 1. The closer the value of AUC is to 1, the higher the accuracy of "the contribution value of each display behavior generated by the machine learning model". It can be seen from the above data that the contribution value of each display behavior obtained by the method provided in this application is the most accurate.

[0172] exist Figures 1a to 7On the basis of the corresponding embodiments, in order to better implement the above solutions of the embodiments of the present application, the following also provides related devices for implementing the above solutions. Figure 8 , Figure 8 A structural schematic diagram of a data processing device provided in an embodiment of the present application, the data processing device 800 includes: a processing module 801, which is used to input a first training sample into a first machine learning model to obtain first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates the probability of at least one first display behavior being interacted; the processing module 801 is used to process the first training sample through a second machine learning model to obtain second information, the second information indicates the sum of contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to at least one first display behavior being interacted; a training module 802 is used to train the second machine learning model according to a loss function until a convergence condition is met, and obtain a second machine learning model that has performed a training operation, wherein the loss function indicates the similarity between the first information and the second information, and the second machine learning model that has performed a training operation is used to obtain the contribution value of each display behavior in at least one display behavior.

[0173] Optionally, the target display behavior is any one of at least one first display behavior, wherein the contribution value of the target display behavior is obtained based on a first value and a second value corresponding to the target display behavior, the first value corresponding to the target display behavior represents the contribution value of the target display behavior itself, and the second value corresponding to the target display behavior represents the influence value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample on the contribution value of the target display behavior.

[0174] Optionally, the first training sample includes the display time of the first display behavior, wherein a contribution value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample to the target display behavior is related to a first time distance, and the first time distance is the distance between the display time of the target display behavior and the display time of other first display behaviors.

[0175] Optionally, the first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

[0176] Optionally, the values ​​of the parameters in the second machine learning model that has been trained are used to obtain a contribution value of each display behavior in at least one display behavior.

[0177] Optionally, the first machine learning model is pre-trained, and the first machine learning model is any one of the following: a recurrent neural network, a convolutional neural network, a fully connected neural network, or an attention-based neural network.

[0178] Optionally, the probability of at least one first display behavior being interacted with is any one of the following: the probability of being clicked in at least one first display behavior, the probability of being downloaded in at least one first display behavior, the probability of being purchased in at least one first display behavior, or the probability of being played in at least one first display behavior.

[0179] It should be noted that the information interaction, execution process, etc. between the modules / units in the data processing device 800 are the same as those in the present application. Figures 1a to 6 The corresponding method embodiments are based on the same concept. For specific contents, please refer to the description in the method embodiments shown above in this application, which will not be repeated here.

[0180] Please continue reading Fig. 9 , Fig. 9 A structural schematic diagram of a contribution value acquisition device provided in an embodiment of the present application, the contribution value acquisition device 900 includes: an acquisition module 901, used to acquire a first sample, the first sample indicates at least one display behavior, and at least one of the display behaviors has been interacted; a generation module 902, used to generate a contribution value of each display behavior pointed to by the first sample through a third machine learning model, the contribution value of each display behavior represents the contribution of each display behavior to the at least one display behavior being interacted, and the third machine learning model is obtained based on a second machine learning model that has performed a training operation; wherein the loss function used in the training operation indicates the similarity between the first information and the second information corresponding to the first training sample, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of at least one first display behavior being interacted, the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the at least one first display behavior being interacted.

[0181] Optionally, the third display behavior is any one of at least one display behavior indicated by the first sample, wherein the contribution value of the third display behavior is obtained based on a first value and a second value corresponding to the third display behavior, the first value corresponding to the third display behavior represents the contribution value of the third display behavior itself, and the second value corresponding to the third display behavior represents the influence value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample on the contribution value of the third display behavior.

[0182] Optionally, the first sample also includes the display time of the display behavior indicated by the first sample, wherein a contribution value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample to the third display behavior is related to a second time distance, and the second time distance is the distance between the display time of the third display behavior and the display time of other display behaviors.

[0183] Optionally, the first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

[0184] It should be noted that the information interaction, execution process, etc. between the modules / units in the contribution value acquisition device 900 are the same as those in the present application. Figure 7 The corresponding method embodiments are based on the same concept. For specific contents, please refer to the description in the method embodiments shown above in this application, which will not be repeated here.

[0185] Next, an execution device provided by an embodiment of the present application is introduced. Fig.10 , Fig.10 A schematic diagram of the structure of an execution device provided in an embodiment of the present application, specifically, the execution device 1000 includes: a receiver 1001, a transmitter 1002, a processor 1003 and a memory 1004 (wherein the number of processors 1003 in the execution device 1000 can be one or more, Fig.10 In the example of FIG. 1003 , the processor 1003 may include an application processor 10031 and a communication processor 10032. In some embodiments of the present application, the receiver 1001, the transmitter 1002, the processor 1003 and the memory 1004 may be connected via a bus or other means.

[0186] The memory 1004 may include a read-only memory and a random access memory, and provides instructions and data to the processor 1003. A portion of the memory 1004 may also include a non-volatile random access memory (NVRAM). The memory 1004 stores processor and operation instructions, executable modules or data structures, or subsets thereof, or extended sets thereof, wherein the operation instructions may include various operation instructions for implementing various operations.

[0187] The processor 1003 controls the operation of the execution device. In a specific application, the various components of the execution device are coupled together through a bus system, wherein the bus system includes not only a data bus but also a power bus, a control bus, and a status signal bus, etc. However, for the sake of clarity, various buses are referred to as bus systems in the figure.

[0188] The method disclosed in the above embodiment of the present application can be applied to the processor 1003, or implemented by the processor 1003. The processor 1003 can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 1003. The above processor 1003 can be a general processor, a digital signal processor (digital signal processing, DSP), a microprocessor or a microcontroller, and can further include an application specific integrated circuit (application specific integrated circuit, ASIC), a field programmable gate array (field-programmable gate array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The processor 1003 can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiment of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor are combined to execute. The software module may be located in a storage medium mature in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 1004, and the processor 1003 reads the information in the memory 1004 and completes the steps of the above method in combination with its hardware.

[0189] The receiver 1001 can be used to receive input digital or character information and generate signal input related to the relevant settings and function control of the execution device. The transmitter 1002 can be used to output digital or character information through the first interface; the transmitter 1002 can also be used to send instructions to the disk group through the first interface to modify the data in the disk group; the transmitter 1002 can also include a display device such as a display screen.

[0190] In the embodiment of the present application, the processor 1003 is used to execute Figure 7 The method for obtaining the contribution value executed by the execution device in the corresponding embodiment. It should be noted that the specific manner in which the application processor 10031 in the processor 1003 executes the above steps is the same as that in the present application. Figure 7The corresponding method embodiments are based on the same concept, and the technical effects they bring are the same as those in this application. Figure 7 The corresponding method embodiments are the same. For specific contents, please refer to the description in the method embodiments shown above in this application, which will not be repeated here.

[0191] The present application also provides a training device. Fig.11 , Fig.11 This is a structural diagram of a training device provided in an embodiment of the present application. Specifically, the training device 1100 is implemented by one or more servers. The training device 1100 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 1122 (for example, one or more processors) and memory 1132, and one or more storage media 1130 (for example, one or more massive storage devices) storing application programs 1142 or data 1144. Among them, the memory 1132 and the storage medium 1130 can be short-term storage or permanent storage. The program stored in the storage medium 1130 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the training device. Furthermore, the central processing unit 1122 can be configured to communicate with the storage medium 1130 to execute a series of instruction operations in the storage medium 1130 on the training device 1100.

[0192] The training device 1100 may also include one or more power supplies 1126, one or more wired or wireless network interfaces 1150, one or more input and output interfaces 1158, and / or, one or more operating systems 1141, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0193] In the embodiment of the present application, the central processor 1122 is used to execute Figures 3 to 6 The data processing method executed by the training device in the corresponding embodiment, the aforementioned training device includes a first training device and / or a second training device. It should be noted that the specific manner in which the central processor 1122 executes the above steps is the same as that in the present application. Figures 3 to 6 The corresponding method embodiments are based on the same concept, and the technical effects they bring are the same as those in this application. Figures 3 to 6 The corresponding method embodiments are the same. For specific contents, please refer to the description in the method embodiments shown above in this application, which will not be repeated here.

[0194] The present application also provides a computer-readable storage medium in which a program for signal processing is stored. When the program is run on a computer, the computer executes the above-mentioned Figures 3 to 6 The steps performed by the first training device and / or the second training device in the method described in the embodiment shown, or the steps performed by the computer as described above Figure 7 The illustrated embodiment describes the steps performed by the device in the method.

[0195] The present application also provides a computer program product which, when executed on a computer, enables the computer to execute the above Figures 3 to 6 The steps performed by the first training device and / or the second training device in the method described in the embodiment shown, or the steps performed by the computer as described above Figure 7 The illustrated embodiment describes the steps performed by the device in the method.

[0196] The execution device and training device provided in the embodiment of the present application may be a chip, which includes: a processing unit and a communication unit. The processing unit may be, for example, a processor, and the communication unit may be, for example, an input / output interface, a pin or a circuit. The processing unit may execute the computer execution instructions stored in the storage unit to enable the chip to execute the above Figures 3 to 6 The data processing method described in the embodiment shown in the figure, or, to enable the chip to execute the above Figure 7 The method for obtaining the contribution value described in the illustrated embodiment. Optionally, the storage unit is a storage unit in the chip, such as a register, a cache, etc. The storage unit may also be a storage unit located outside the chip in the wireless access device, such as a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM), etc.

[0197] For details, please refer to Fig.12 , Fig.12 A schematic diagram of the structure of a chip provided in an embodiment of the present application, the chip can be expressed as a neural network processor NPU 120, NPU 120 is mounted on the host CPU (Host CPU) as a coprocessor, and the host CPU assigns tasks. The core part of the NPU is the operation circuit 1203, which is controlled by the controller 1204 to extract matrix data from the memory and perform multiplication operations.

[0198] In some implementations, the operation circuit 1203 includes multiple processing units (Process Engine, PE) inside. In some implementations, the operation circuit 1203 is a two-dimensional systolic array. The operation circuit 1203 can also be a one-dimensional systolic array or other electronic circuits capable of performing mathematical operations such as multiplication and addition. In some implementations, the operation circuit 1203 is a general-purpose matrix processor.

[0199] For example, assume there is an input matrix A, a weight matrix B, and an output matrix C. The operation circuit takes the corresponding data of matrix B from the weight memory 1202 and caches it on each PE in the operation circuit. The operation circuit takes the matrix A data from the input memory 1201 and performs matrix operation with matrix B, and the partial result or final result of the matrix is ​​stored in the accumulator 1208.

[0200] The unified memory 1206 is used to store input data and output data. The weight data is directly transferred to the weight memory 1202 through the direct memory access controller (DMAC) 1205. The input data is also transferred to the unified memory 1206 through the DMAC.

[0201] BIU stands for Bus Interface Unit, i.e., bus interface unit 1210 , which is used for the interaction between AXI bus, DMAC and instruction fetch buffer (IFB) 1209 .

[0202] The bus interface unit 1210 (Bus Interface Unit, BIU for short) is used for the instruction fetch memory 1209 to obtain instructions from the external memory, and is also used for the storage unit access controller 1205 to obtain the original data of the input matrix A or the weight matrix B from the external memory.

[0203] DMAC is mainly used to transfer input data in the external memory DDR to the unified memory 1206 or to transfer weight data to the weight memory 1202 or to transfer input data to the input memory 1201.

[0204] The vector calculation unit 1207 includes multiple operation processing units, and further processes the output of the operation circuit when necessary, such as vector multiplication, vector addition, exponential operation, logarithmic operation, size comparison, etc. It is mainly used for non-convolutional / fully connected layer network calculations in neural networks, such as Batch Normalization, pixel-level summation, upsampling of feature planes, etc.

[0205] In some implementations, the vector calculation unit 1207 can store the processed output vector to the unified memory 1206. For example, the vector calculation unit 1207 can apply a linear function and / or a nonlinear function to the output of the operation circuit 1203, such as linear interpolation of the feature plane extracted by the convolution layer, and then, for example, a vector of accumulated values ​​to generate an activation value. In some implementations, the vector calculation unit 1207 generates a normalized value, a pixel-level summed value, or both. In some implementations, the processed output vector can be used as an activation input to the operation circuit 1203, for example, for use in a subsequent layer in a neural network.

[0206] An instruction fetch buffer 1209 connected to the controller 1204 is used to store instructions used by the controller 1204;

[0207] Unified memory 1206, input memory 1201, weight memory 1202 and instruction fetch memory 1209 are all on-chip memories. External memories are private to the NPU hardware architecture.

[0208] in, Figures 3 to 7 The operations of each layer in each machine learning model shown can be performed by the operation circuit 1203 or the vector calculation unit 1207.

[0209] The processor mentioned in any of the above places may be a general-purpose central processing unit, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the program of the above-mentioned first aspect method.

[0210] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0211] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0212] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0213] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A data processing method, characterized in that: The method comprises: Inputting a first training sample into a first machine learning model to obtain first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates a probability of the at least one first display behavior being interacted with; Processing the first training sample by a second machine learning model to obtain second information, wherein the second information indicates the sum of contribution values ​​of each first display behavior in the at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the at least one first display behavior being interacted with; The second machine learning model is trained according to the loss function until a convergence condition is met, thereby obtaining the second machine learning model that has performed a training operation, wherein the loss function indicates the similarity between the first information and the second information, and the second machine learning model that has performed a training operation is used to obtain a contribution value of each display behavior in at least one display behavior.

2. The method according to claim 1, characterized in that: The target display behavior is any one of the at least one first display behavior, wherein the contribution value of the target display behavior is obtained based on a first value and a second value corresponding to the target display behavior, the first value corresponding to the target display behavior represents the contribution value of the target display behavior itself, and the second value corresponding to the target display behavior represents the influence value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample on the contribution value of the target display behavior.

3. The method according to claim 2, characterized in that The first training sample includes the display time of the first display behavior, wherein a contribution value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample to the target display behavior is related to a first time distance, and the first time distance is the distance between the display time of the target display behavior and the display time of the other first display behaviors.

4. The method according to any one of claims 1 to 3, characterized in that: The first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, and the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

5. The method according to any one of claims 1 to 3, characterized in that: The values ​​of the parameters in the second machine learning model that has performed the training operation are used to obtain the contribution value of each display behavior in at least one display behavior.

6. The method according to any one of claims 1 to 3, characterized in that: The first machine learning model is pre-trained, and the first machine learning model is any one of the following: a recurrent neural network, a convolutional neural network, a fully connected neural network, or an attention-based neural network.

7. The method according to any one of claims 1 to 3, characterized in that: The probability of the at least one first display behavior being interacted with is any one of the following: the probability of being clicked in the at least one first display behavior, the probability of being downloaded in the at least one first display behavior, the probability of being purchased in the at least one first display behavior, or the probability of being played in the at least one first display behavior.

8. A method for obtaining a contribution value, characterized in that: The method comprises: Acquire a first sample, where the first sample indicates at least one display behavior, and the at least one display behavior has been interacted with; generating a contribution value of each display behavior pointed to by the first sample through a third machine learning model, wherein the contribution value of each display behavior represents a contribution made by each display behavior to the at least one display behavior being interacted with, and the third machine learning model is obtained based on the second machine learning model that has performed a training operation; Among them, the loss function adopted in the training operation indicates the similarity between the first information and the second information corresponding to the first training sample, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of the at least one first display behavior being interacted, and the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in the at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of the at least one first display behavior.

9. The method according to claim 8, characterized in that The third display behavior is any one of the at least one display behavior indicated by the first sample, wherein the contribution value of the third display behavior is obtained based on the first value and the second value corresponding to the third display behavior, the first value corresponding to the third display behavior represents the contribution value of the third display behavior itself, and the second value corresponding to the third display behavior represents the influence value of other display behaviors other than the third display behavior in the at least one display behavior indicated by the first sample on the contribution value of the third display behavior.

10. The method according to claim 9, characterized in that The first sample also includes the display time of the display behavior indicated by the first sample, wherein the contribution value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample to the third display behavior is related to the second time distance, and the second time distance is the distance between the display time of the third display behavior and the display time of the other display behaviors.

11. The method according to any one of claims 8 to 10, characterized in that: The first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, and the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

12. A data processing device, characterized in that: The device comprises: a processing module, configured to input a first training sample into a first machine learning model to obtain first information output by the first machine learning model, wherein the first training sample indicates at least one first display behavior, and the first information indicates a probability of the at least one first display behavior being interacted with; The processing module is configured to process the first training sample through a second machine learning model to obtain second information, wherein the second information indicates the sum of contribution values ​​of each first display behavior in the at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the at least one first display behavior being interacted with; A training module is used to train the second machine learning model according to a loss function until a convergence condition is met, thereby obtaining the second machine learning model that has performed a training operation, wherein the loss function indicates the similarity between the first information and the second information, and the second machine learning model that has performed a training operation is used to obtain a contribution value of each display behavior in at least one display behavior.

13. The device according to claim 12, characterized in that The target display behavior is any one of the at least one first display behavior, wherein the contribution value of the target display behavior is obtained based on a first value and a second value corresponding to the target display behavior, the first value corresponding to the target display behavior represents the contribution value of the target display behavior itself, and the second value corresponding to the target display behavior represents the influence value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample on the contribution value of the target display behavior.

14. The device according to claim 13, characterized in that The first training sample includes the display time of the first display behavior, wherein a contribution value of other first display behaviors other than the target display behavior in at least one first display behavior indicated by the first training sample to the target display behavior is related to a first time distance, and the first time distance is the distance between the display time of the target display behavior and the display time of the other first display behaviors.

15. The device according to any one of claims 12 to 14, characterized in that The first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, and the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

16. The device according to any one of claims 12 to 14, characterized in that The values ​​of the parameters in the second machine learning model that has performed the training operation are used to obtain the contribution value of each display behavior in at least one display behavior.

17. The device according to any one of claims 12 to 14, characterized in that The first machine learning model is pre-trained, and the first machine learning model is any one of the following: a recurrent neural network, a convolutional neural network, a fully connected neural network, or an attention-based neural network.

18. The device according to any one of claims 12 to 14, characterized in that The probability of the at least one first display behavior being interacted with is any one of the following: the probability of being clicked in the at least one first display behavior, the probability of being downloaded in the at least one first display behavior, the probability of being purchased in the at least one first display behavior, or the probability of being played in the at least one first display behavior.

19. A device for obtaining contribution value, characterized in that: The device comprises: An acquisition module, configured to acquire a first sample, wherein the first sample indicates at least one display behavior, and the at least one display behavior has been interacted with; a generating module, configured to generate a contribution value of each display behavior pointed to by the first sample through a third machine learning model, wherein the contribution value of each display behavior represents a contribution made by each display behavior to the at least one display behavior being interacted with, and the third machine learning model is obtained based on the second machine learning model that has performed a training operation; Among them, the loss function adopted in the training operation indicates the similarity between the first information and the second information corresponding to the first training sample, the first training sample indicates at least one first display behavior, the first information is obtained through the first machine learning model, the first information indicates the probability of the at least one first display behavior being interacted, and the second information is obtained through the second machine learning model, the second information indicates the sum of the contribution values ​​of each first display behavior in the at least one first display behavior, and the contribution value of each first display behavior represents the contribution of each first display behavior to the interaction of the at least one first display behavior.

20. The device according to claim 19, characterized in that The third display behavior is any one of the at least one display behavior indicated by the first sample, wherein the contribution value of the third display behavior is obtained based on the first value and the second value corresponding to the third display behavior, the first value corresponding to the third display behavior represents the contribution value of the third display behavior itself, and the second value corresponding to the third display behavior represents the influence value of other display behaviors other than the third display behavior in the at least one display behavior indicated by the first sample on the contribution value of the third display behavior.

21. The device according to claim 20, characterized in that The first sample also includes the display time of the display behavior indicated by the first sample, wherein the contribution value of other display behaviors other than the third display behavior in at least one display behavior indicated by the first sample to the third display behavior is related to the second time distance, and the second time distance is the distance between the display time of the third display behavior and the display time of the other display behaviors.

22. The device according to any one of claims 19 to 21, characterized in that The first training sample is derived from a training data set, and there is a second training sample in the training data set, wherein the second training sample is obtained based on a third training sample, and the third training sample includes a display time of at least one second display behavior, and at least one display time in the third training sample is updated to obtain the second training sample.

23. An execution device, characterized in that: comprising a processor and a memory, the processor being coupled to the memory, The memory is used to store programs; The processor is configured to execute the program in the memory so that the execution device executes the method according to any one of claims 8 to 11.

24. A training device, characterized in that comprising a processor and a memory, the processor being coupled to the memory, The memory is used to store programs; The processor is used to execute the program in the memory so that the training device performs the method as claimed in any one of claims 1 to 7.

25. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, and when the program is executed on a computer, the computer is caused to execute the method according to any one of claims 1 to 11.

26. A computer program product, characterized in that The computer program product comprises a program, and when the program is run on a computer, the computer is caused to execute the method according to any one of claims 1 to 11.